Mr J. M. Macfarlane on Lcpidophloios. 185 



slight variation in form having been reckoned sufficient for 

 specific distinction." Let us see how this applies in the 

 plant before us. The leaf-cushions at their bases spread 

 out and are attached to the stem so as to enclose a quadri- 

 lateral area, in the centre of which the impression of the 

 rhomboidal cicatrix may appear. This 1 believe to be 

 what Dawson figures as Lepidophloios tetragonus, while 

 Schimper's Lepidodendron quadratum is, I doubt not, the 

 same without the cicatrix. The sub-epidermal part of the 

 bark is marked only by the elongated prominences which 

 the leaf-bundles give rise to on passing out into the leaves. 

 The fact already mentioned, viz., that the leaf-cushions 

 imbricate in a downward direction is important, and will 

 assist me in my future generalisations. Schimper seems 

 to be the only writer who has pointed this out hitherto. 

 That such does take place has been clearly demonstrated 

 by my finding branches in the act of dichotomising ; more- 

 over, the leaf-cushions follow a course exactly the opposite 

 of that pursued in the forking of the stem. From this it 

 might be supposed that the leaves did not closely surround 

 the stem as in Ulodendron, but that they spread out in a 

 lax fashion, as has been shown by Corda. In all my 

 specimens wliere dichotomous division is seen, the angle 

 which the two branches form is almost exactly 90°, a 

 divergence very much greater than most Lepidodendra, 

 where it is only about 45°. Such then is the general 

 structure of stems from 2 inches to 1 foot in diameter. 

 When we pass to those of smaller size, we find that two 

 forms are encountered, which are essentially alike in the 

 nature and disposition of the leaf- cushions, the form of the 

 scars, the number of vascular markings on these, the angle 

 of dichotomy, and the final branch termination in a rather 

 abrupt manner, but which differ in two important re- 

 spects, since in the one a remarkable series of tubercles 

 gradually appear, arranged in quincuntial order, as also 

 the passage, according to Williamson, from the medul- 

 lary vascular cylinder of a number of vascular bundles 

 proceeding out into these tubercles. Both these peculi- 

 arities are wanting in the second. In other words, I 

 regard Lepidophloios as undoubtedly terminating in two 

 forms of branch, the one being what we have hitherto 



